To support EFSA in the preparatory work for the assessment of Tolerable Upper Intake Levels (UL) for vitamin B6, a tailored ‘high level’ protocol was developed, based on a template protocol and taking into account the specificities of vitamin B6. This protocol outlined the methods, the relevant endpoints and priority adverse health effects in relation to high intakes of vitamin B6. Systematic reviews (SR) were conducted following a tailored literature search, data extraction, evidence appraisal (i.e. risk of bias (RoB) assessment) and evidence synthesis. Narrative reviews (NR) were conducted to gather contextual evidence relevant to the interpretation of the main body of evidence. In the SR on the dose‐response relationship between vitamin B6 and peripheral neuropathy, 3793 records were identified and reduced to 32 individual records after screening for eligibility. The evidence appraisal revealed an overall moderate to high RoB of the individual studies and an overall rating of the total body of evidence as very low due mainly due to the type of data from case studies/case reports and limited number of other data. The available evidence confirmed a high degree of inter‐individual variability to sensitivity to a high exposure of vitamin B6 in relation to development of peripheral neuropathic outcomes but did not allow to determine a No Observed Adverse Effect Level or Lowest Observed Adverse Effect Level. In the SR on developmental toxicity, 4941 records were identified and reduced to 23 individual records including human and animal studies after screening for eligibility. The available data demonstrated a high degree of heterogeneity with respect to exposure and adverse health outcomes and overall, showed no positive or causal relationship between vitamin B6 intake and adverse developmental effects, including congenital defects. The NRs showed a paucity of data on high intake of vitamin B6 vitamers and their metabolism.

BACKGROUND The global shift towards plant-based diets has been increasing, with more people making the transition for various reasons. In vulnerable subgroups such as patients or older adults, the transition to plant-based diets deserves attention due to the potentially detrimental consequences of lower protein quantity and quality. Guidance in meal formulation to assure optimal protein quantity and quality is needed. This guidance is a complex and time-consuming task for consumers but a viable task for artificial intelligence. OBJECTIVE We aimed to develop a digital tool that uses artificial intelligence to ensure sufficient protein quality in plant-based meals for older adults, utilizing a co-creation design approach that involved interdisciplinary collaboration. METHODS Together with Dutch and Danish dietitians and older adults as potential end-users, three focus groups were held to identify their preferences and dealbreakers. Focus group sessions were based on a User-Task-Environment analysis, the Walt-Disney method, the Brainwriting method, and a Cognitive Walkthrough. The front-end and back-end development of the tool took place in parallel and well-synced to the focus group sessions. RESULTS Both dietitians and older adults from Dutch and Danish sites expressed high interest in a tool that provides feedback and background information on protein quality, sustainability, and (micro)nutrients. The User-Task-Environment analysis delivered useful details on the user, the task of the tool, and the environment in which the tool will be used. The Walt-Disney method delivered usable and realistic solutions to the four challenges presented. Thirty-two percent of the solutions on all themes presented with the Brainwriting method appeared to be highly feasible and relevant, and these suggestions were integrated into the tool. The Cognitive Walkthrough identified certain screens as unclear, necessitating revisions for improved understandability. CONCLUSIONS The results of the focus groups revealed the need for a tool in the guidance of meal planning to ensure sufficient protein quality. A first co-created prototype of the tool is ready, which will be further developed and optimized according to the needs of end users. This user-driven tool, designed to guide meal composition for sufficient protein quality, holds promise for enhancing older adults' dietary choices and health outcomes as they transition to plant-based diets.